Running head: SPATIAL MEMORY

Spatial Memory Following Temporal Lobe Resection

Catherine M. Diaz-Asper1, Stephen Dopkins2, Samuel J. Potolicchio, Jr.3, Anthony Caputy4

George Washington University

Washington, DC

1 Clinical Brain Disorders Branch, NIMH, NIH 2 Dept. of Psychology, The George Washington University 3 Dept. of Neurology, The George Washington University 4 Dept. of Neurosurgery, The George Washington University Please address correspondence to: Stephen Dopkins Department of Psychology 2125 G Street, NW Washington, DC 20052 Tel: (202) 994-6315 Fax: (202) 994-1602 E-mail: dopkins@gwu.edu

2

Abstract

The present study sought a clearer understanding of spatial memory function consequent to

temporal lobe resection, and, in particular, of spatial memory function with respect to two- as well as

three-dimensional frames of reference. Relative to a group of 15 control participants, a group of 15

epilepsy patients with right temporal resections demonstrated deficits of memory for locations in a two-

dimensional display. A group of 13 epilepsy patients with left temporal resections did not demonstrate

such deficits. The right and the left resection groups both demonstrated deficits of memory for item-

location relationships in a two-dimensional display. The right but not the left resection group

demonstrated deficits of memory for item-location relationships in a three-dimensional display. The

differing results that were observed for item-location relationships in two- and three-dimensional displays

were attributed to differences in the way item information is bound with location information concerning

two- and three-dimensional domains.

KEYWORDS: spatial memory, visual memory, medial temporal lobe, temporal lobe epilepsy

3

Spatial memory following temporal lobe resection

There is widespread agreement that structures within the human temporal lobes play a critical role

in memory function. The initial evidence for this view emerged from studies of the patient H.M. After

undergoing bilateral resection of the medial temporal lobe, H.M. was found to have severe deficits of

long-term declarative memory (Miller, Corkin, & Tueber, 1968; Sagar, Cohen , Corkin, & Growdon,

1985; Scoville & Milner, 1957).

H.M.’s memory deficit has subsequently been modeled fairly successfully in non-human primates

(Gaffan, 1994; Murray & Mishkin, 1986; Ridley, Timothy, MacLean, & Baker, 1995; Suzuli, Zola-

Morgan, Squire, & Amaral 1993; Zola, Squire, Teng, Stefanacci, & Clark, 2000). Across delays ranging

from seconds to minutes, monkeys with similar medial temporal ablations have demonstrated deficits on

visual and tactual non-matching- to-sample tasks and on matching-to- location tasks (Squire, 1992; Zola-

Morgan & Squire, 1985). Further work with primates has attempted to localize memory function more

precisely. For example, deficits on simple non-matching-to-sample tasks have been demonstrated in

monkeys with damage to the perirhinal and entorhinal cortex (Murray & Mishkin, 1986; Suzuli et al.,

1993). In contrast, less severe impairments have been demonstrated on such tasks in monkeys with

damage restricted to the hippocampus [Zola et al., 2000).

At the same time, human neuropsychological research has sought to identify the substrates of

different sorts of long-term declarative memory within the medial temporal region. Early work explored

the lateralization of such memory. Dysfunction of the left medial temporal lobe was associated with

deficits of verbal memory (Blakemore & Falconer, 1967; Kimura, 1963; Pillon, Bazin, Deweer, Ehlre,

Baulac, & Dubois, 1999). Dysfunction of the right medial temporal lobe was associated with deficits of

nonverbal memory, including memory for visual and spatial information (Milner, 1968, Pigott & Milner,

1993; Pillon et al., 1999; Rains & Milner, 1994; Smith & Milner, 1981; Smith & Milner, 1989).

4

More recently, efforts have been made toward a fuller articulation of the effects of human medial

temporal dysfunction on long-term nonverbal memory. Few definitive conclusions have been reached,

however. One complicating factor is that many different facets of nonverbal memory can be

distinguished. Within spatial memory, for example, a distinction can be drawn on the basis of the

underlying reference frame, which can be either two- or three-dimensional (Stepankova, Fenton,

Pastalkova, Kalina, & Bohbot, 2004). An independent distinction can be drawn between location

memory, in which one simply remembers a particular location in a spatial domain, and item-location

memory, in which one remembers the location of a particular item in a domain (Rains & Milner, 1994;

Smith & Milner, 1981; Smith & Milner, 1989). Within three-dimensional spatial memory, a further

distinction can be drawn between egocentric memory, in which locations are remembered with reference

to one’s own location, and allocentric memory, in which locations are remembered without reference to

one’s own location (Feigenbaum & Morris, 2004; Goldstein, Canavan, & Polkey, 1989; Holdstock,

Mayes, Cezayirli, Aggleton, & Roberts, 1999; Incisa della Roccchetta, Samson, Ehrle, Denos, Hasboun,

& Baulac, 2004; O’Keefe & Nadel, 1978).

The present study attempted a relatively comprehensive and analytical examination of spatial

memory consequent to medial-temporal dysfunction. In particular, the object was to examine two- as well

as three-dimensional spatial memory. This was deemed to be important for two reasons. First, the bulk of

previous work in spatial memory has focused on memory either for three-dimensional displays or for

representations of three-dimensional scenes (Goldstein et al., 1989; Nunn, Graydon, Polkey, & Morris,

1999; Nunn, Polkey, & Morris, 1998; Pigott & Milner, 1993; Rains & Milner, 1994; Smith & Milner,

1981; Smith & Milner, 1989). Consequently, little is known about the effects of temporal lobe

dysfunction on two-dimensional spatial memory. As the results of the present study show, two- and three-

dimensional spatial memory do not always follow the same pattern. Second, imaging studies have

5

recently begun to explore two-dimensional spatial memory (Johnsrude, Owen, Crane, Milner, & Evans,

1999; Moscovitch, Kapur, Kohler, & Houle, 1995; Owen, Milner, Petrides, & Evans, 1996). It is

important to seek convergence between neuropsychological and imaging data.

Three groups of participants were tested: a group that had undergone resection of the anterior right

temporal lobe (RTLR), a group that had undergone resection of the anterior left temporal lobe (LTLR),

and a group of healthy adults. Portions of the hippocampus as well as the parahippocampal gyrus had

been resected in the RTLR and LTLR groups.

Experiment 1 examined two-dimensional spatial memory, making a distinction between location

and item-location memory. To establish a context for understanding item-location memory, the

experiment also tested two-dimensional visual memory for item information. Experiment 2 examined

three-dimensional item-location memory, making a distinction between egocentric and allocentric

memory.

The study differed in several methodological respects from many previous studies in this area.

First, the stimulus items for the tests of item and item-location memory were abstract visual figures rather

than objects. In this respect, the study followed Nunn et al. (1998) rather than several other previous

studies (Nunn et al., 1999; Owen et al., 1996; Smith & Milner, 1981; Smith & Milner, 1989). The

rationale here was that the use of abstract stimuli would limit the involvement of verbal processes in

performance of the tasks, and clarify the implications of the results with respect to spatial memory.

Second, the different tasks in each experiment were matched for difficulty on the basis of pilot

data. Thus, it was expected that the control participants would not differ in performance on the tasks for a

given experiment. The object was to assess the deficiency of the two TLR groups relative to the control

participants on each task. This is in contrast to past studies, which have either made no attempt to match

tasks for difficulty (Barr, 1997; Goldstein et al., 1989; Hermann, Seidenberg, Wyler, & Haltiner, 1993;

6

Owen et al., 1996; Smith & Milner, 1981; Smith & Milner, 1989) or, alternatively, have used a group-

matching procedure in which TLR and control participants are matched for performance on tasks of one

sort and then deficits are demonstrated in the TLR participants on tasks of another sort (Nunn et al.,

1998; Nunn et al., 1999). The virtues of the present procedure are twofold. First, it avoids problems

associated with the group-matching procedure, namely that TLR and control participants typically have to

be tested under different conditions (e.g. different retention intervals) in order to achieve equivalent levels

of performance on the task on which they are matched. It is conceivable that different sorts of memory

may be called into play under these different conditions (e.g. different sorts of memory may be involved

at different retention intervals). The second virtue of the present procedure is that it allows for more

efficient use of the experimental data. In particular contrast to the group matching procedure (Nunn et al.,

1998; Nunn et al., 1999), it is possible under the present procedure to assess the performance of the RTLR

and LTLR participants relative to the control participants on all tasks for a given experiment.

Experiment 1

This experiment tested the two-dimensional spatial memory of RTLR, LTLR, and control

participants, making a distinction between location and item-location memory. To aid in understanding the results

of the item-location memory test, the experiment also tested two-dimensional visual memory for item information.

The object was to discover whether the RTLR and/or LTLR participants differed from the control

participants on any of these forms of memory. Each of the experimental tasks consisted of an exposure

phase, in which a sequence of target stimuli was presented, and a test phase, in which the participant

attempted to distinguish the target stimuli from distractor stimuli.

The item task tested two-dimensional visual memory for the sorts of information used in

identifying items. Participants were required to remember the identities of abstract visual figures,

presented on a computer screen. The location task tested two-dimensional location memory. Participants

were required to remember locations on a computer screen. The item-location task tested two-dimensional

7

item-location memory. Participants were required to remember the locations of abstract visual figures on

a computer screen.

Method

Participants

Temporal lobe resection participants. These were drawn from a group of 43 patients who had

undergone a selective unilateral resection of the anterior temporal lobe as therapy for intractable epilepsy

(Weibe, Blume, Girvin, & Eliasziw, 2001). The seizure foci for the patients were idenitified with ictal

video-EEG monitoring. The extent of the resection for each patient was determined by intra-operative

EEG measurement. An initial portion of the anterior- lateral temporal lobe was first defined by

electrocortographic mapping. Following the resection of this portion, the electrocorticography array was

again placed on the temporal lobe. Recordings were obtained from the uncut surface of the temporal lobe,

from the resection surface, and from the hippocampus. Stimulation was again employed, including direct

stimulation of the hippocampus. If an electrically-active focus was defined, further cortical resection was

undertaken until the area was quiescent from an eleptogenic standpoint. The medial temporal structures,

including the amygdala and a portion of the hippocampus (3.0-cm) were then resected using

microsurgical techniques. The fornix was spared.

Because age at seizure onset, disease duration, seizure frequency (Dodrill & Matthews, 1992), and

anti-epileptic drug treatment (Devinsky, 1995), are known to influence memory abilities in temporal lobe

epileptics, participants in the two TLR groups were matched on these factors. The resulting sample

contained 31 patients (16 RTLR and 15 LTLR). One RTLR patient withdrew from the study after

completing one task, and two LTLR patients were excluded after being tested, due to the presence of

potentially confounding neurological conditions (cerebral palsy and spina bifida). Thus, twenty-eight

patients were included in the final analyses (15 RTLR and 13 LTLR). Two of the LTLR participants had

8

superior quadrant deficits. Because these participants achieved the required criterion on the

perceptual/attentional screening task (to be described later), they were retained in the study.

Surgeon’s notes indicated that the final group of RTLR participants had undergone resections of

the inferior, middle, and superior gyri, averaging, respectively, 3.29, 2.89, and 1.50 cm from the temporal

pole, with standard deviations of .34, 1.05, and 1.43 cm, respectively. The LTLR participants had

undergone resections averaging, respectively, 2.95, 2.55, and 1.05 cm, with standard deviations of .71,

1.16, and .99 cm, respectively. Notice that more cortex was resected from the inferior than the superior

gyrus; thus the resection involved an oblique cut extending from anterior/rostral to posterior/caudal.

Notice, also, that the amounts of cortex resected from the right and left TLR participants were more

nearly equivalent than would have been the case under some other resection procedures. In fact, the

amounts of tissue resected from the three cortical areas did not differ in the RTLR and LTLR groups

(inferior: F(1,22) = 1.43, MSe = .458; middle: F(1,22) = 1.93, MSe = .355; superior: F(1,22) < 1. These

and all other statistical tests were conducted against an alpha value of .05.). Analysis of post-surgery

imaging data for the final group of participants indicated that the procedure removed 25-40% of the

temporal lobe, including the amygdala and approximately 40% of the hippocampus and the surrounding

parahippocampal gyrus. Pathology reports indicated the presence of hippocampal atrophy or sclerosis for

9 of the RTLR and 9 of the LTLR participants. Tumors were resected from two of RTLR participants. All

TLR participants were tested at a minimum of 5 months following their surgery. All were left-hemisphere

dominant for speech.

Healthy control participants. These were drawn from a pool of 19 healthy volunteers. Criteria for

inclusion in the pool included the stated absence of any potentially confounding medical or neurological

conditions (e.g., history of head injury, serious cardiovascular disease, or stroke), and the stated absence

of drug and alcohol abuse. Six of the 19 participants in the pool were relatives or friends of TLR

9

participants. The remaining 13 participants were recruited from the surrounding community. From this

pool, a sample of 15 individuals was selected that were matched to the patient groups in terms of age, sex,

handedness, years of education, ethnicity, and estimated intelligence level.

Intelligence (IQ) was assessed for all participants using the Information and Picture Completion

subtests of the Wechsler Adult Intelligence Scale- Revised (WAIS-R) (Weschler, 1981). A composite

score was obtained from these subtests with an age-adjusted formula created by Kaufman, Ishikuma, and

Kaufman-Packer (1991) (see also Boone (1992)). Strength of handedness was ascertained using the

Edinburgh Handedness Inventory (Oldfield, 1971). No participant (TLR or control) had an estimated IQ

lower than 80, or an age less than 16 or greater than 60. All participants provided written informed

consent prior to undertaking the experiments.

Demographic information for the three participant groups is presented in Table 1. The groups did

not differ in terms of age (F(2,40)=1.01), sex (Χ2(2,N=43)=2.11), estimated IQ (F(2,40)=.48), handedness

(F(2,40)=.86), or education level (F(2,40)=.09). The two TLR groups did not differ in terms of age at

seizure onset (F(1,26)=.13), duration of epilepsy (F(1,26)=.29), current seizure frequency (F(1,26)=2.20),

or number of anti-epileptic medications taken (F(1,26)=.08).

Insert Table 1 here

Materials

Each target and distractor display for the item task consisted of an abstract visual figure in the

center of the computer screen. The figures were constructed, in the manner of Musen and Treisman

(1990) and Blaxton and Theodore (1997), by connecting points in a 3 X 3 array of points. To construct a

given figure, five pairs of points, chosen at random, were connected, with the constraints that (a) the

resulting line segments formed a connected figure, (b) no line segment was repeated in either direction,

and (c) a line did not simultaneously cross a column and a row in the point array. Any repetitions of

10

figures, or highly verbalizable figures (e.g., common shapes or alphanumeric characters) were eliminated

through visual inspection. The array points were erased for the final presentation of the figures. Each

figure was approximately 2 cm X 2 cm in size.

Each target and distractor display for the location task consisted of a 2 cm X 2 cm white square at

a certain location within a region that was defined on the computer screen by a continuous irregular

border. The border remained constant across all displays. The rationale for including this border depended

on the distinction that has been proposed between the categorical and the coordinate modes of spatial

processing. According to this proposal, relationships such as “above,” “beside,” etc. are assessed with

categorical processing, whereas quantitative measures of distance are assessed with coordinate

processing. Categorical processing is based in the left hemisphere, whereas coordinate processing is based

in the right hemisphere (Kosslyn, 1987; Kosslyn, 1994). The border was included to encourage nonverbal

coordinate involvement in the location task. It was reasoned that participants would be encouraged to

locate the white square in a given display in terms of a set of distances from points on the border.

Each target and distractor display for the item-location task consisted of an abstract visual figure,

of the same sort used for the item task, at a certain location within the same irregular-bordered region that

was used for the location task (See Figure 1).

Insert Figure 1 here

Because the testing phase for each task required participants to emit positive responses to target

displays and negative responses to distractors, the stimulus set for each task was composed of pairs of

displays, with one member of each pair serving as the target, and the other as the distractor. For the item

task, the members of each target/distractor (T/D) pair were visually similar (with at least 3 out of 5 line

segments in common). For the location task, the members of each T/D pair resembled one another in the

placement of the white square. The position of the square differed by no less than 1 cm but no more than 3

11

cm. For the item-location task, a distractor was formed for each target by placing the visual figure of the

target in a position previously occupied by the visual figure of a different target during the exposure

phase. The displays for the exposure and test phases of the three tasks were presented in the same quasi-

random order to all participants.

Task matching. The tasks for the experiment were matched for difficulty using the procedure of

Chapman and Chapman (1978). The procedure involved pilot testing two parallel sets of 15 T/D pairs for

each task. Hence, 90 T/D pairs were tested overall. Participants comprised a convenient sample of 136

undergraduate students, who received class credit for their participation. Sixty-three participants piloted

one set of T/D pairs for all three tasks, and 67 piloted a second set. The testing procedure was the same as

is later described for the experiment proper. When pilot testing was completed, the corrected recognition

(hits – false alarms) score for each T/D pair was calculated. Fifteen T/D pairs were then selected for the

final version of each task, with the corrected recognition score for each selected pair coming from the

middle of the distribution for its set of T/D pairs. Different visual figures were used in the pairs for the

three tasks. The 15 T/D pairs that were selected for the item, location, and item-location tasks had mean

corrected recognition scores of .29, .31, and .32, respectively. The Kuder-Richardson formula 20 (a

coefficient of internal consistency that is used for items scored dichotomously) was then computed for the

T/D pairs that had been selected for each task. The Kuder-Richardson scores for the item, location, and

item-location tasks were .75, .84, and .93, respectively. Finally, an additional 6 participants were tested on

the T/D pairs that had been selected for the three tasks. In this sample, the mean corrected recognition

scores for the three tasks were .31, .34, and .35. The Kuder-Richardson scores were .76, .86, and .93.

Procedure

Perceptual/attentional screening. Although recent work suggests that temporal lobectomy patients

do not suffer from simple deficits of perception and attention, all participants were screened to ensure

12

adequate perceptual and attentional capacities prior to undertaking the experimental tasks (Mendola,

Rizzo, Cosgrove, Cole, Black, & Corkin, 1999). The screening task was computer-based and included

components that tested perception and attention with respect to item and location information. All

participants completed the item component first. This consisted of 30 trials (excluding two practice trials

that served to familiarize participants with the task). At the beginning of each trial, a visual figure (of the

same sort as was used in the experiment) was presented for 150 ms (Hellige & Michimata, 1989; Rybash

& Hoyer, 1992). Following a 1-second delay, a second figure was presented for 150 ms. Participants were

required to indicate, as quickly as possible, whether the second figure matched the first. The location

component was similar, except that the visual figure was presented at a certain location within the region

that was defined in the center of the computer screen by the irregular border. Participants were required

to determine whether the second presentation of the figure occurred in the same location as the first. The

delay interval between the two presentations in each trial was kept deliberately brief (1 sec) to ensure that

perceptual and attentional, rather than mnemonic, processes were engaged (Hellige & Michimata, 1989).

On the basis of findings with a pilot sample, participants were required to achieve at least 25 out of a

possible 30 correct responses on each component of the screening task in order to be included in the

study. Both of the LTLR participants with superior quadrant deficits attained this level of performance.

One TLR participant was excluded for failure to attain this level of performance.

Overview of experiment proper. Each of the three tasks consisted of an exposure phase and a test

phase. The test phase for each task followed its associated exposure phase (e.g., item exposure phase,

followed by item test phase; location exposure phase, followed by location test phase, etc.). Previous

studies assessing long-term memory function have used varying retention intervals. The delay between

the exposure and the test phase has ranged from minutes to weeks (Abrahams, Pickering, Polkey, &

Morris, 1997; Smith & Milner, 1981). Previous studies have also varied in whether or not the participant

13

was actively engaged in a “filler” task between exposure and test. The present experiment used a three-

minute, filled, retention interval. The filler task consisted of questions pertaining to the participant’s

medical or social background. The presentation order of the three tasks was counterbalanced across

participants.

Exposure phase. The exposure phase for the item task consisted of 15 trials, on each of which a

different visual figure was presented in the center of the computer screen (without the irregular border).

Each trial lasted for 5 seconds, and was followed by a 3-second inter-trial interva l. Participants were

instructed to remember the figures for later testing. The exposure phase for the location task consisted of

15 trials, on each of which a white square was presented in a different location within the region that an

irregular border defined on the computer screen. The white square and the irregular border were the same

for all trials. Each trial lasted for 5 seconds and was followed by a 3-second inter-trial interval.

Participants were instructed to remember the indicated locations for later testing. The exposure phase for

the item-location task consisted of 15 trials, on each of which a different visual figure was presented in a

different location within the irregular-bordered region. Each trial lasted for 5 seconds, and was followed

by a 3-second inter-trial interval. Participants were instructed to remember the figures and their locations

for later testing.

Test phase. The test phase for each of the three tasks commenced immediately following the three-

minute filled delay. Participants received both verbal and on-screen instructions immediately prior to each

testing phase. For the item task, 30 displays (15 targets and 15 distractors) were presented, in sequence.

Each display consisted of a visual figure in the center of the computer screen (without the irregular

border). Participants indicated whether the figure had been presented during the exposure phase. For the

location task, 30 displays (15 targets and 15 distractors) were presented. Each display consisted of the

white square in a certain location within the irregular-bordered region. Participants indicated whether or

14

not the square had been presented in the indicated location during the exposure phase. For the item-

location task, 30 displays (15 targets and 15 distractors) were presented. Each display consisted of a visual

figure in a certain location within the irregular-bordered region. Participants indicated whether the figure

had been presented in the indicated location. For all three tasks, participants pressed the “Y” and the “N”

key to make their responses.

Participants also completed a de-briefing questionnaire regarding their performance in the three

tasks. The questionnaire asked participants to describe the mnemonic strategies that they had used in

performing each task. For the item task, participants were asked whether they had been reminded of

familiar things when studying the figures, whether they had formulated names or verbal descriptions for

the figures, whether they had used some other strategy, or whether they had used no particular strategy.

For the location task, participants were asked whether they had remembered the location of the white

square relative to the contour of the irregular border, whether they had remembered the location relative

to the shape of the computer screen, whether they had remembered the location in some other way, or

whether they had used no particular strategy. For the item-location task, participants were asked whether

they had used a combination of the strategies indicated earlier, whether they had used some other strategy,

or whether they had used no particular strategy.

Results

Memory Tests

Figure 2 illustrates the mean corrected recognition score for each of the three tasks as a function of

group membership. A series of one-way ANOVA’s revealed significant group differences on the Item

task (F(2,40)=8.96, MSe=.049), the Location task (F(2,40)=10.21, MSe=.020), and the Item-Location task

(F(2,40)=4.52, MSe=.048).

Insert Figure 2 here

15

Planned comparisons revealed significant differences between the RTLR participants and the

control participants on the item task (F(1,28) = 19.97, MSe = .043, Partial Eta Squared = .42), the location

task (F(1,28) = 21.42, MSe = .015, Partial Eta Squared = .43), and the item-location task (F(1,28) = 7.38,

MSe = .05, Partial Eta Squared = .21). In addition, significant differences were revealed between the

LTLR participants and the control participants on the item (F(1,26) = 4.76, MSe = .05, Partial Eta

Squared = .16) and the item-location tasks (F(1,26) = 5.31, MSe = .049, Partial Eta Squared = .17), and

between the RTLR and the LTLR participants on the location task (F(1,26) = 13.88, MSe = .018, Partial

Eta Squared = .35).

Ninety-five percent confidence intervals were examined for the corrected recognition data from

each group to ascertain the likelihood that memory capacity was completely absent, with the result that

the data reflected random performance. For the LTLR and the control participants, confidence intervals

ranged from 0.47 to 8.20 across the three tasks. As a result, it is reasonably certain that the observed data

do not reflect random performance. The same can be said for the performance of the RTLR participants

on the item-location task (CI = .16 to 3.58). However, the performance of the RTLR participants on the

item (CI = -.40 to 3.07) and the location (CI = -.10 to 2.10) tasks suggests that performance may have

been random here.

Although the differences were not significant, the three participant groups did nonetheless differ in

estimated IQ and distribution of males and females. Therefore, one-way ANCOVAs were calculated for

the three groups on the three tasks, with sex and estimated IQ as covariates, to ascertain whether these

variables affected performance. The main effect of the covariates was not significant for the location and

the item-location tasks, but there was a significant effect of estimated IQ (F(1,38)=4.68, MSe=.041) and

sex (F(1,38)=4.83, MSe=.041) on item task performance. After controlling for the effects of estimated IQ

and sex, the main effect of group on the item task remained highly significant (F(2,38)=9.66, MSe=.041).

16

Pairwise comparisons revealed the same pattern as reported above: The RTLR (F(1,26) = 19.63, MSe =

.041, Partial Eta Squared = .34) and the LTLR (F(1,24) = 8.86, MSe = .036, Partial Eta Squared = .27)

participants both performed significantly more poorly than the control participants on this task.

A multivariate repeated-measures ANOVA confirmed that the control group’s performance was

matched across the three experimental tasks (F(2,13)=1.63).

Mnemonic Strategies

In order to determine whether the groups differed in the strategies used for memorizing the target

stimuli, chi-square analyses were performed on the results of the de-briefing questionnaire. On the item

task, a majority of the participants in each group (11/13 LTLR, 13/15 RTLR, and 11/15 control

participants) stated that they were reminded of familiar things when studying the figures. The groups did

not differ in their use of this strategy (Χ2 (2,N=43)=1.01). On the location task, a majority of the

participants in each group (12/13 LTLR, 14/15 RTLR, and 14/15 control) stated that they remembered the

location of the white square relative to the contour of the irregular border. Again, the groups did not differ

in their use of this strategy (Χ2 (2,N=43)=.02). On the item-location task, a majority of the participants in

each group (8/13 LTLR, 11/15 RTLR, and 11/15 control) used the former two strategies in combination.

Again, the three groups did not differ in their choice of strategy for this task (Χ2 (2,N=43)=4.72).

Memory Test Performance and Size of Resection

The surgeon’s notes for the TLR participants listed measurements across three regions of the

temporal lobe: the inferior gyrus, the middle gyrus, and the superior gyrus. Analyses were conduced for

14 of the 15 RTLR participants to assess the correlation between measures of performance on the three

tasks and amounts of tissue resected from the three regions. (The surgeon’s note for one RTLR participant

provided non-specific resection information). A constant amount of hippocampal tissue was removed in

each resection, so this structure was not considered in the analysis. None of the correlations between the

17

task variables and the resection measurements was significant at the .05 level. The same analyses were

also conducted for 10 of the 13 LTLR participants, revealing no significant relationships. The results of

the correlational analyses are presented in Table 2.

Insert Table 2 here

Discussion

None of the tasks was more difficult than any other for the control participants. By implication,

the preliminary task-matching process was successful.

The RTLR and LTLR participants both performed more poorly than the control participants on the

Item task. These results are somewhat unexpected, given that the majority of previous reports have linked

visual memory deficits with right temporal dysfunction (Barr, 1997; Jones-Gotman, Zatorre, Olivier,

Andermann, Cendes, Staunton, McMackin, Siegel, & Wieser, 1997; Milner, 1968). The results are not

unprecedented, however. Several previous reports have linked visual memory deficits with left temporal

dysfunction or with left and right dysfunction (Hornak, Oxbury, Oxbury, Iverson, & Gaffan, 1997;

Maguire & Cipolotti, 1998). Future work will be necessary to disentangle the factors that control visual

memory deficits in temporal lobe patients.

The results for the item task imply that the information needed to identify the stimulus items was

bilaterally represented. Several conceptions of this bilateral representation are possible. On one hand, the

items might have been represented differently in the two hemispheres – in the verbal modality in the left

hemisphere and in the visual modality in the right hemisphere. On the other hand, the stimuli might have

been represented in the same way in the two hemispheres - in terms that were neither verbal nor visual but

that instead might be called amodal. Several points can be made against the joint verbal-visual

conception. First, it goes against the principle of parsimony. Second, the visual figures used in the present

study were designed to be resistant to verbal coding. Third, a majority of participants in all of the groups

18

indicated on the de-briefing questionnaire that they performed the item task by noting resemblances

between the figures and familiar things but not by formulating names or verbal descriptions for the

figures. For example participants may have called up a conceptual representation of a thing that a given

figure resembled. Because the representation existed at the conceptual level it would have reflected visual,

auditory, olfactory, and tactile aspects of the object. In general, then, the results are consistent with the

view that the item memory process represented the figures in amodal terms.

The RTLR participants performed more poorly than the LTLR and the control participants on the

location task. These results reinforce previous reports of a location memory deficit in RTLR patients. The

results constitute useful new information, however. Whereas most previous studies have shown deficits of

three-dimensional location memory in RTLR patients (O’Reilly & Rudy, 2001), the present results show

deficits of two-dimensional location memory. As subsequent results will show, two- and three-

dimensional spatial memory do not always show the same pattern.

The RTLR and LTLR participants performed more poorly than the control participants on the

item-location task. These results again reinforce previous reports of an item-location memory deficit in

RTLR patients. Again, however, the results constitute new information. Whereas most previous studies

have shown deficits of three-dimensional item-location memory (Goldstein et al., 1989; Nunn et al., 1998;

Nunn et al., 1999; Smith & Milner, 1981; Smith & Milner, 1989), the present results show deficits of two-

dimensional item-location memory. Furthermore, whereas previous RTLR deficits of three-dimensional

item-location memory have occurred in the context of intact LTLR performance, the present RTLR

deficits of two-dimensional item-location memory occur in the context of similar LTLR deficits.

In interpreting the results for the item-location task, we must bear in mind that the item and

location tasks, on one hand, and the item-location task, on the other hand, made independent demands

upon participants. Whereas the item and location tasks required participants to remember the figures and

19

locations that had been presented, the item-location task provided participants with the figures and

locations that had been presented and required them to remember the way the figures and locations had

been linked. Thus, the item-location task tapped the binding of items and locations (Chalfonte & Johnson,

1996; O’Reilly & Rudy, 2001; Shimamura, 2002). What, then, are the implications of these results?

Notably, the information that bound items and locations was bilaterally recorded. Thus, the pattern was

the same as for item information and different from the pattern for location information. This issue will be

discussed in greater detail later.

Experiment 2

This experiment tested the three-dimensional item-location memory of the RTLR, LTLR, and

control participants, making a distinction between egocentric and allocentric item-location memory. The

object was to discover whether the RTLR and/or LTLR participants differed from the control participants

in either egocentric or allocentric three-dimensional item-location memory and to compare the degree to

which the RTLR and/or LTLR participants differed from the controls in the two sorts of memory.

During the exposure phase of the egocentric item-location task, a series of displays was presented,

with each display consisting of an abstract two-dimensional visual figure in a certain location on a

tabletop surface. During the test phase, participants distinguished displays that had been presented during

the exposure phase from new displays in which old figures were presented in new locations. The

procedure for the allocentric item-location task was the same as for the egocentric item-location task,

except that the displays were rotated either 90, 180, or 270 degrees between the exposure and the test

phase. Goldstein et al. (1989) used a similar procedure to test egocentric and allocentric item-location

memory.

Method

20

Materials

Each target and distractor display was constructed in the same way as for the item-location task of

Experiment 1. The displays were printed in black ink on sheets of 8.5”X11” paper, approximately the

same size as in Experiment 1. The sheets on which the displays were printed were contained within a 3”

ring-binder. Separate sheets of green paper were placed between the display sheets for masking purposes.

The 15 T/D pairs for both the egocentric and allocentric item-location tasks were formed in the

same way as the T/D pairs for the item-location task of Experiment 1. For the test phase of the allocentric

item-location task, the sheets on which the target displays were printed were rotated clockwise in

increments of 90°. (Notice that the target figures remained in the same orientation relative to the irregular

border, because the whole sheet was rotated). Nine sheets were rotated 90° at test, 11 were rotated 180°,

and 10 were rotated 270°. The displays for the exposure and test phases of the two tasks were presented

in the same quasi-random order to all participants.

Task matching. This was carried out as for Experiment 1. Two parallel sets of 15 T/D pairs were

pilot tested for each task. Hence, 60 T/D pairs were tested overall. The participants comprised a

convenient sample of 139 undergraduate students, who received class credit for their participation. Fifty-

two participants piloted the egocentric item-location task with one set of stimuli and the allocentric item-

location task with another set of stimuli. A further 52 participants piloted the two tasks in the opposite

manner. The testing procedure was the same as is later described for the experiment proper. The 15 T/D

pairs that were initially selected for the egocentric and allocentric item-location tasks had mean corrected

recognition scores of .18 and .19, respectively. Because accuracy was rather low, alterations were made to

the targets and distractors to make them more discriminable (the visual figures were moved closer to the

border). The altered T/D pairs for the two tasks were subsequently tested again with a different sample of

35 participants. The altered T/D pairs for the egocentric and allocentric item-location tasks achieved mean

21

corrected recognition scores of .29 and .30, respectively. The Kuder-Richardson scores for the egocentric

and allocentric item-location pairs were .88 and .92, respectively. Different visual figures were used in the

displays for the two tasks.

Procedure

Overview. The testing phase for each task followed the exposure phase for that task, with a three-

minute interval inserted between exposure and test. This interval was usually filled with unrelated

conversation. The presentation order of the tasks was counterbalanced across participants.

Exposure phase. The exposure phase for both the egocentric and the allocentric item-location tasks

consisted of 15 trials, on each of which the loose- leaf binder was laid on the table, directly in front of the

participant, open to the display sheet for the current trial. The participant saw the sheet for 5 seconds and

was instructed to remember the location of the figure relative to the irregular border. Each trial was

followed by a 3-second inter-trial interval.

Test phase. The test phase for both of the tasks commenced immediately following the three-

minute filled delay. Thirty displays (15 previously-seen targets and 15 distractors), were presented on the

table, in sequence, directly in front of the participant. For the egocentric item-location task, participants

were required to indicate whether or not each figure had appeared previously in the indicated location.

For the allocentric item-location task, participants were told that the page had been rotated from its

orientation at exposure and required to indicate whether or not each figure had appeared in the indicated

location relative to the irregular border. For both tasks, participants indicated their responses by writing

“Y” or “N” on pre-printed answer sheets.

Results

Memory Tests

Figure 3 shows the mean corrected recognition scores for the two tasks as a function of group

22

membership. One-way ANOVA’s failed to reveal significant group differences on the egocentric item-

location task (F(2,40)=1.78, MSe=.024, p = .182). There was a trend towards group differences on the

allocentric item-location task (F(2,40)=3.02, MSe=.041, p=.06). A post-hoc comparison revealed that the

performance of the RTLR group on the egocentric and allocentric item-location tasks was significantly

poorer than the performance of the control group on these tasks (F(1,28) = 8.64, MSe = .032, Partial Eta

Squared = .24), and that the difference between the LTLR and the control participants did not differ on

the two tasks (F(1,28) < 1).

Insert Figure 3 here

Ninety-five percent confidence intervals were examined for the corrected recognition data for each

group, to ascertain the likelihood that the data reflect random performance. Across all three participant

groups, CIs were > 0 for the egocentric item-location task, suggesting that the observed data do not reflect

random performance. The same can be said for the LTLR and control participants on the allocentric item-

location task (CIs ranged from .67 to 5.46). However, the RTLR group’s performance on the allocentric

item-location task suggests that performance may have been random here (CI = -.46 to 2.73).

One-way ANCOVAs were also conducted, with estimated IQ and gender as covariates. No

significant main effects of the covariates were observed for either the egocentric or the allocentric item-

location tasks.

A multivariate repeated-measures ANOVA confirmed that the control group’s performance was

matched across the two experimental tasks, (F(1,14)=3.16).

Task Performance and Size of Resection

Measures of performance on both tasks were correlated with the resection measurements for 14 of

the 15 RTLR participants. None of the correlations between the task variables and the resection variables

was significant at the .05 level. The same correlations were also conducted for 10 of the 13 LTLR

23

participants, revealing no significant relationships. The results of the correlational analyses are presented

in Table 3.

Insert Table 3 here

Discussion

None of the tasks was more difficult than any other for the control participants. The preliminary

task-matching process was evidently successful.

In comparison to the control participants, the RTLR participants performed significantly more

poorly on the egocentric and allocentric item-location tasks. These results reinforce previous reports of a

three-dimensional item-location deficit in RTLR patients (Nunn et al., 1998; Nunn et al., 1999; Smith &

Milner, 1981; Smith & Milner, 1989). Furthermore, as in previous work, the RTLR deficits occurred in

the context of intact LTLR performance.

The RTLR participants were equally deficient on the tests of egocentric and allocentric item-

location memory. These results contrast with previous reports that RTLR participants are relatively more

deficient on tests of allocentric than egocentric item-location memory (Gaffan, 1994).

The results of the present experiment offer an interesting contrast to the results of Experiment 1. In

Experiment 1, both RTLR and LTLR participants showed deficits of two-dimensional item-location

memory. In the present experiment, RTLR but not LTLR participants showed deficits of three-

dimensional item-location memory. In both experiments, the item-location tasks tapped information on

the binding of item and location information. The stimulus items and the procedures for testing item-

location memory were the same in the two experiments. The primary difference between the two

experiments was that the stimulus items were located in a two-dimensional domain in Experiment 1 and a

three-dimensional domain in Experiment 2. The contrast between the results of the two experiments will

be addressed below.

24

General Discussion

The present results contribute to an emerging picture of spatial memory consequent to temporal

lobe dysfunction. These results reinforce earlier reports regarding three-dimensional spatial memory.

More importantly, the results augment the extant picture with new information regarding two-dimensional

spatial memory. In Experiment 1, RTLR but not LRLR patients showed deficits of two-dimensional

location memory. At the same time, both RTLR and LTLR patients showed deficits of two-dimensional

item-location memory. The exploration of two-dimensional spatial memory consequent to temporal lobe

resection is particularly useful in light of the importance of the two-dimensional reference frame in

imaging studies of spatial memory (Johnsrude et al., 1999; Moscovitch et al., 1995; Owen et al, 1996).

In fleshing out the picture of two-dimensional spatial memory consequent to temporal lobe

dysfunction, the present results suggest complexities that have been heretofore unnoted. Of greatest

interest, in this respect, are the results for the two- and three-dimensional item-location tasks. Whereas the

RTLR and LTLR participants both showed deficits in the two-dimensional task, only the RTLR

participants showed deficits in the three-dimensional task. Although the two- and three- dimensional

item-location tasks were not explicitly matched for difficulty, the two tasks were quite similar except for

the difference in dimensionality. Thus, the differences that were observed across tasks may require an

explanation.

In attempting to explain these differences, it is useful to consider the item-location tasks in the

context of the item task. The RTLR and LTLR participants both showed deficits in the latter task. Thus,

across participant groups, performance in the two- but not the three-dimensional item-location task

followed the same pattern as performance in the item task. Why was this the case? It is worth

remembering that the two-dimensional item-location task involved a simpler form of space than the three-

dimensional item-location task. The space of the two-dimensional task paralleled the space of the retina.

25

In contrast, the space of the three-dimensional task had to be inferred from the two-dimensional retinal

image on the basis of depth cues. In light of this difference between the two item-location tasks, one

possible explanation of the observed pattern runs as follows: Item information was dominant in the

memory representation that was created to bind item and two-dimensional location information. At the

core of a given representation was the item in question. The item’s location was recorded as an added

piece of information associated with the item. In contrast, location information was dominant in the

memory representation that was created to bind item and three-dimensional location information. At the

core of a given representation was the location in question. The presence of the item at that location was

recorded as an added piece of information associated with the location. The difference between the two

cases arose because three-dimensional location is more complex than two-dimensional location. An

item’s three-dimensional location was too complex to be recorded as an added piece of information.

This proposed explanation can accommodate the present results as follows: Consider, first, the fact

that RTLR and LTLR participants both showed deficits of two-dimensional item-location memory. These

deficits derived from deficits of item memory; the binding of item and two-dimensional location

information depended on the representation of item information. Consider, second, the fact that RTLR but

not LTLR participants showed deficits of three-dimensional item-location memory. We assume that

RTLR but not LTLR participants had deficits of three-dimensional location memory. Although this was

not shown in the present study, it has been shown in previous work (Rybash & Hoyer, 1992). The

assumption derives further support from the fact that RTLR but not LTLR participants showed deficits of

two-dimensional location memory in the present study. Given this assumption, the deficits that RTLR but

not LTLR participants showed in three-dimensional item-location memory derived from deficits of three-

dimensional location memory; the binding of item and three-dimensional location information depended

on the representation of location information.

26

The proposed explanation is consistent with results from imaging studies. Imaging results

comport, in particular, with the idea that the binding of item and two-dimensional location information

depends on the representation of item information. Imaging studies have shown the entorhinal cortex to be

activated during item memory tasks (Fried, Cameron, Yashar, Fong, & Morrow, 2002). (Results from

non-human primates provide a similar picture; if non-matching and matching- to-sample are taken as

analogs of recognition, deficits in item recognition have been associated with ablation of the entorhinal

cortex (Murray, 2000). Imaging studies have also shown the entorhinal cortex to be activated during two-

dimensional item-location memory tasks (Owen et al., 1996). The present results are consistent with the

implication that the same brain area is involved in memory for item and two-dimensional item-location

memory. In addition, the present results are consistent with the view that the entorhinal cortex is the

particular brain area involved, inasmuch as this area was included in the resected portion of the temporal

lobe for both the RTLR and LTLR groups.

Of course, the proposed explanation remains purely speculative at this point and will need to be

borne out by further results. Regardless of the ultimate fate of this particular explanation, the results that

are demonstrated here for two-dimensional item-location memory will require a future accounting.

The present study also produced results that are inconsistent in several respects with previous

work in nonverbal memory. The results bear, first, on the question of the relative deficiency of visual and

spatial memory consequent to right medial temporal dysfunction. Conflicting positions have been taken

with respect to this question, with Barr (1997) and Nunn et al. (1998) arguing that right medial temporal

dysfunction produces larger deficits of visual and spatial memory, respectively. The present results

support neither of these previous positions; in Experiment 1, the RLTR participants were equally deficient

relative to the control participants on the tests of visual memory (the item task) and spatial memory (the

location and item-location tasks). The present results also bear on the question of the relative deficiency

27

of egocentric and allocentric memory consequent to right medial temporal dysfunction. In previous work,

it has been argued that allocentric memory is more deficient than egocentric memory (Goldstein et al.,

1989). The present results do not support this view; in Experiment 2, the RLTR participants were equally

deficient relative to the control participants on the tests of egocentric and allocentric item-location

memory. In considering the differences between the present and previous results, it is important remember

the measures that were taken here to match the tasks of the two experiments for difficulty. In light of these

measures, the present results may deserve more credence than previous findings.

The present results are also of some interest with respect to visual memory. In contrast to many

previous studies, in which visual memory deficits have been reported only for RTLR patients, both RTLR

and LTLR patients showed deficits of visual memory, as indexed by the item memory task. Further work

will be necessary to sort out the factors that determine visual memory deficits in temporal lobe patients.

Finally, the present results have implications for the pre- and post-surgical assessment of patients

undergoing temporal lobe resection. These results suggest that the pattern of visual-spatial deficits

following temporal lobe resection is complex. Ideally, pre- and post-assessments should tap item,

location, and item-location memory. Location and item-location memory should ideally be assessed in

two as well as three dimensions. In this context, it should be noted that an informal survey of memory

problems among the patient participants revealed fewer complaints in the RTLR than the LTLR group.

Several possible limitations of the study should be noted. First, the sample may have been too

small. Assuming the effect size of Nunn et al. (1998), a sample of 15 RTLR and 15 LTLR patients, and a

critical value of .05, pre-experimental calculations placed the power of the experimental tasks at .8. This

estimate may have been overly optimistic, however. A larger sample would have increased the power of

the tasks to detect differences among the groups. Second, the level of performance on the tests of the

28

study may have been low enough among the control participants as to cause floor effects among the

patient participants.

In conclusion, the present study demonstrated deficits for RTLR and LTLR patients on tests of

item memory and two-dimensional item-location memory. The study demonstrated deficits for RTLR but

not LTLR patients on tests of two-dimensional location memory and three-dimensional item-location

memory.

29

References

Abrahams, S., Pickering, A., Polkey, C., & Morris, R. (1997). Spatial memory deficits in patients

with unilateral damage to the right hippocampal formation. Neuropsychologia, 35, 11-24.

Barr, W. (1997). Examining the right temporal lobe’s role in nonverbal memory. Brain and

Cognition, 35, 26-41.

Blakemore, C. B., & Falconer, M. A. (1967). Long-term effects of anterior temporal lobectomy on

certain cognitive functions. Journal of Neurology, Neurosurgery, and Psychiatry, 30, 364-367.

Blaxton, T., & Theodore, W. (1997). The role of the temporal lobes in recognizing visuospatial

material: Remembering versus knowing. Brain and Cognition, 35, 5-25.

Boone, D. E. (1992). Evaluation of Kaufman’s short forms of the WAIS-R with psychiatric

inpatients. Journal of Clinical Psychology, 48, 239-245.

Chalfonte, B., & Johnson, M. K. (1996). Feature memory and binding in young and older adults.

Memory & Cognition, 24, 403-416.

Chapman, L. J., & Chapman, J. P. (1978). The measurement of differential deficit. Journal of

Psychiatric Research, 14, 303-311.

Devinsky, O. (1995). Cognitive and behavioral effects of antiepileptic drugs. Epilepsia, 36, 46-65.

Dodrill, C. B., & Matthews, C. G. (1992). The role of neuropsychology in the assessment and

treatment of persons with epilepsy. American Psychologist, 47, 1139-1142.

Feigenbaum, J. D., & Morris, R. G. (2004). Allocentric versus egocentric spatial memory after

unilateral temporal lobectomy in humans. Neuropsychology, 18, 462-472.

Fried, I., Cameron, K. A., Yashar, S., Fong, R., & Morrow, J. W. (2002). Inhibitory and excitatory

responses of single neurons in the human medial temporal lobe during recognition of faces and objects.

Cerebral Cortex, 12, 575-84.

30

Gaffan, D. (1994). Scene-specific memory for objects: A model of episodic memory impairment

in monkeys with fornix transection. Journal of Cognitive Neuroscience, 6, 305-320.

Goldstein, L. H., Canavan, A. G. M., & Polkey, C. E. (1989). Cognitive mapping after unilateral

temporal lobectomy. Neuropsychologia, 27, 167-177.

Haist, F., Bowden, G. J., & Mao, H. (2001). Consolidation of human memory over decades

revealed by functional magnetic imaging. Nature Neuroscience, 4, 1139-45.

Hellige, J. B., & Michimata, C. (1989). Categorization versus distance: Hemispheric differences

for processing spatial information. Memory and Cognition, 17, 770-776.

Hermann, B., Seidenberg, M., Wyler, A., & Haltiner, A. (1993). Dissociation of object

recognition and spatial localization abilities following temporal lobe lesions in humans.

Neuropsychology, 7, 343-350.

Holdstock, J. S., Mayes, A. R., Cezayirli, E., Aggleton, J. P., & Roberts, N. (1999). A comparison

of egocentric and allocentric spatial memory in medial temporal lobe and Korsakoff amnesics. Cortex, 35,

479-501.

Hornak, J., Oxbury, S., Oxbury, J., Iversen, S. D., & Gaffan, D. (1997). Hemifield-specific visual

recognition memory impairments in patients with unilateral temporal lobe removals. Neuropsychologia,

35, 1311-1315.

Incisa della Rocchetta, A., Samson, S, Ehrle, N., Denos, M., Hasboun, D., & Baulac, M. (2004).

Memory for visuospatial location following selective hippocampal sclerosis. Neuropsychology, 18, 15-28.

Jones-Gotman, M., Zatorre, R. J., Olivier, A., Andermann, F., Cendes, F., Staunton, H.,

McMackin, D., Siegel, A. M., & Wieser, H.-G. (1997). Learning and retention of words and designs

following excision from medial or lateral temporal- lobe structures. Neuropsychologia, 35, 963-973.

31

Johnsrude, I. S., Owen, A. M., Crane, J., Milner, B. & Evans, A. C. (1999). A cognitive activation

study of memory for spatial relationships. Neuropsychologia, 37, 829-841.

Kaufman, A. S., Ishikuma, T., & Kaufman-Packer, J. L. (1991). Amazingly short forms

of the WAIS-R. Journal of Psychoeducational Assessment, 9, 4-15.

Kimura, D. (1963). Right temporal lobe damage. Archives of Neurology, 8, 264-71.

Kosslyn, S. M. (1987). Seeing and imagining in the cerebral hemispheres: A computational

approach. Psychological Review, 94, 148-175.

Kosslyn, S. M. (1994). Image and brain: The resolution of the imagery debate. Cambridge, MA:

MIT Press.

Maguire, E. A., & Cipolotti, L. (1998). Selective sparing of topographic memory. Journal of

Neurology, Neurosurgery & Psychiatry, 65, 903-909.

Mendola, J. D., Rizzo, J. F., Cosgrove, G. R., Cole, A. J., Black, P, & Corkin, S. (1999). Visual

discrimination after anterior temporal lobectomy in humans. Neurology, 52, 1028-1037.

Milner, B. (1968). Visual recognition and recall after right temporal- lobe excision.

Neuropsychologia, 6, 191-209.

Milner, B., Corkin, S., & Teuber, H. L. (1968). Further analysis of the hippocampal amnesic

syndrome: 14-year followup study of H.M. Neuropsychologia, 6, 215-234.

Moscovitch, M., Kapur, S., Kohler, S., & Houle, S. (1995). Distinct neural correlates of visual

long-term memory for spatial location and object identity: A PET study in humans. Proceedings of the

National Academy of Science, 92, 3721-3725.

Murray, E. A. (2000). Memory for objects in nonhuman primates. In M. S. Gazzaniga (Ed.), The

New Cognitive Neurosciences, pp. 753-763. Cambridge, MA: MIT Press.

32

Murray, E. A., & Mishkin, M. (1986). Visual recognition in monkeys following rhinal cortical

ablations combined with either amydalectomy or hippocampectomy. The Journal of Neuroscience, 6,

1991-2003.

Musen, G., & Treisman, A. (1990). Implicit and explicit memory for visual patterns. Journal of

Experimental Psychology: Learning, Memory, and Cognition, 16, 127-137.

Nunn, J. A., Graydon, F. J. X., Polkey, C. E., & Morris, R. (1999). Differential spatial memory

impairment after right temporal lobectomy demonstrated using temporal titration. Brain, 122, 47-59.

Nunn, J., Polkey, C., & Morris, R. (1998). Selective spatial memory impairment after right

unilateral temporal lobectomy. Neuropsychologia, 36, 837-848.

O’Keefe, J., & Nadel, L. (1978). The hippocampus as a cognitive map. Oxford: Clarendon Press.

Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh Inventory.

Neuropsychologia, 9, 97-113.

O’Reilly, R. C., & Rudy, J. W. (2001). Conjunctive representations in learning and memory:

Principles of cortical and hippocampal function. Psychological Review, 108, 311-345.

Owen, A. M., Milner, B., Petrides, M., & Evans, A. C. (1996). A specific role for the right

parahippocampal gyrus in the retrieval of object-location: A positron emission

tomography study. Journal of Cognitive Neuroscience, 8, 588-602.

Pigott, S., & Milner, B. (1993). Memory for different aspects of complex visual scenes after

unilateral temporal or frontal- lobe resection. Neuropsychologia, 31, 1-15.

Pillon, B., Bazin, B., Deweer, B., Ehrle, N., Baulac, M., & Dubois, B. (1999). Specificity of

memory deficits after right or left temporal lobectomy. Cortex, 35, 561-571.

Rains, G. D., & Milner, B. (1994). Right-hippocampal contralateral-hand effect in the recall of

spatial location in the tactual modality. Neuropsychologia, 32, 1233-1242.

33

Ridley, R.M., Timothy, C.J., Maclean, C. J., & Baker, H. F. (1995). Conditional learning and

memory impairments following neurotoxic lesion of the CA1 field of the hippocampus. Neuroscience, 67,

263-275.

Rybash, J. M., & Hoyer, W. J. (1992). Hemispheric specialization for categorical and coordinate

Spatial representations: A reappraisal. Memory and Cognition, 20, 271-276.

Sagar, H. H., Cohen, N. J.,Corkin, S. & Growdon, J. M. (1985). Dissociations among processes in

remote memory. In D. S. Olton, E. Gamzu, & S. Corkin (Eds.), Memory Dysfunctions, pp. 533-535. New

York: NY Academy of Science.

Scoville, W., & Milner, B. (1957). Loss of recent memory after bilateral hippocampal lesions.

Journal of Neurology, Neurosurgery, and Psychiatry, 20, 11-21.

Shimamura, A. P. (2002). Relational binding theory and the role of consolidation in memory

retrieval. In L. R. Squire, & D. L. Schacter (Eds.), Neuropsychology of memory (3rd ed.), pp. 61-72.

New York: Guilford Press.

Smith, M., & Milner, B. (1981). The role of the right hippocampus in the recall of spatial

location. Neuropsychologia, 19, 781-793.

Smith, M. L., & Milner, B. (1989). Right hippocampal impairment in the recall of spatial location:

Encoding deficit or rapid forgetting. Neuropsychologia, 27, 71-81.

Squire, L. (1992). Memory and the hippocampus: A synthesis from findings with rats, monkeys,

and humans. Psychological Review, 99, 195-231.

Stepankova, K., Fenton, A, Pastalkova, E., Kalina, M., & Bohbot, V. (2004). Object- location

memory impairment in patients with thermal lesions to the right or left hippocampus. Neuropsychologia,

42, 1017-1028.

34

Suzuki, W. A., Zola-Morgan, S., Squire, L. R., & Amaral, D. G. (1993). Lesions of the perirhinal

and parahippocampal cortices in the monkey produce long- lasting memory impairment in the visual and

tactual modalities. The Journal of Neuroscience, 13, 2430-2451.

Wechsler, D. (1981). WAIS-R manual: Wechsler Adult Intelligence Scale-Revised. San Antonio,

TX: Psychological Corporation.

Wiebe, S., Blume, W. T., Girvin, J. P., & Eliasziw, M. (2001). A randomized, controlled trial of

surgery for temporal- lobe epilepsy. New England Journal of Medicine, 345, 311-318.

Zola, S. M., Squire, L. R., Teng, E., Stefanacci, L., & Clark, R. E. (2000). Impaired recognition

memory in monkeys after damage to limited to the hippocampal region. The Journal of Neuroscience, 20,

451-463.

Zola-Morgan, S. M., & Squire, L. R. (1985). Medial temporal lesions in monkeys impair memory

on a variety of tasks sensitive to human amnesia. Behavioral Neuroscience, 99, 22-34.

35

Figure Captions

Figure 1. Example of the type of stimuli used in the Item-Location task of Experiment 1 and the

Egocentric and Allocentric Item-Location tasks of Experiment 2.

Figure 2. Mean corrected recognition score as a function of participant group on the Item, Location, and

Item-Location tasks of Experiment 1.

Figure 3. Mean corrected recognition score as a function of participant group on the Egocentric and

Allocentric Item-Location tasks of Experiment 2.

36

Table 1

Means and standard deviations of participant characteristics

RTLR LTLR Control

N 15 13 15

Male / female 7/8 3/10 7/8

Edinburgh Handedness Inventory [36] 77.90 (24.70) 74.12 (24.03) 62.81 (44.10)

Age 41.53 (11.89) 43.08 (11.28) 37.60 (8.55)

Years of Education 14.67 (2.50) 14.77 (3.09) 14.40 (1.53)

Estimated IQ 94.89 (13.98) 99.28 (16.91) 98.69 (7.23)

Age at seizure onset 16.83 (9.76) 17.63 (15.02) -

Duration of epilepsy 21.40 (9.59) 20.54 (11.57) -

No. anti-epileptic drugs taken 1.00 (0.78) 1.08 (0.64) -

Current seizure frequency (Engel∇) 1.20 (0.41) 1.46 (0.52) -

Age and time variables reported in years

∇ Engel level 1 = seizure free; 2 = rare seizures; 3 = worthwhile improvement; 4 = no worthwhile

improvement.

37

Table 2

Experiment 1: Pearson Correlations between Amounts of Tissue Resected and Task Variables

Identity Location Identity-Location

RTLR Group

Inferior gyrus -.06 -.02 .05

Middle gyrus .21 .02 -.09

Superior gyrus -.16 -.23 -.19

LTLR Group

Inferior gyrus .21 -.16 -.29

Middle gyrus -.02 -.29 -.46

Superior gyrus -.53 .08 -.38

38

Table 3

Experiment 2: Pearson Correlations between Amounts of Tissue Resected and Task Variables

Egocentric Allocentric

RTLR Group

Inferior gyrus .31 -.46

Middle gyrus .38 -.47

Superior gyrus .35 -.28

LTLR Group

Inferior gyrus .12 .17

Middle gyrus -.02 .25

Superior gyrus -.42 -.38

39

40

41